VMD-L Mailing List

From: Axel Kohlmeyer (akohlmey_at_gmail.com)
Date: Fri Oct 02 2009 - 11:52:09 CDT

On Fri, 2009-10-02 at 01:01 -0700, Ignacio Fernández Galván wrote:

> > > I can generate Gaussian "cube" files, either with Gaussian itself or
> > > with Molden, but those are pre-calculated grid data, and they won't
> >
> > they won't what?
>
> Sorry, it was cut... I mean the cube files won't make any
> use of the Orbital representation: the grid spacing cannot
> be changed, for instance. Cube files work fine with
> Isosurface, however.

right. cube files contain coordinate and volumetric data, so
they cannot work with orbital, as that requires wavefunction
data, which is a new feature.

> > > So I wonder if this new feature of VMD will be really
> > > useful for me.
> >
> > not right now unless you take over writing a plugin to get
> > it done faster than others can do it (our group doesn't do
> > much quantum chemistry, so there is always the chance than
> > some other problem arises that i need to take care of and
> > thus keeps me from returning to work on the gaussian plugin.
> > i have not been able to really touch the code since late
> > spring...)
>
> Hmm... I might have a try at that, but last time I tried I
> had some problems compiling VMD, I'll have a look at
> compiling plugins. Actually I think I would be more

compiling a single plugin for dynamic linkage is straightforward.
you don't really have to use the full VMD build infrastructure
for that. there should be some info in the developer docs online.
i can also send you a makefile that i've been using for the
HOOMD plugin (which is not part of the VMD distribution), that
works well on different kinds of linux machines.

> interested in reading Molden files, which also seem easier
> to parse.

the format looks simpler, but there are some problematic issues
with different kinds of optional sections providing redundant
data and not knowing in advance which will be present and in
what order. that makes parsing molden files almost as nasty
a xml files.

> I've put a sample file at
> <http://djelibeibi.unex.es/files/molden-test.zip>,
> with a double goal: First, to show the structure, which is
> explained in <http://www.cmbi.ru.nl/molden/molden_format.html>
> too, in case someone wants to have a look at it. Second,
> to show that the current Molden plugin does not read this file,
> and one of the reasons seems to be the fact that it has
> "[MOLDEN FORMAT]" instead of "[Molden Format]".

if you look at the changelog of this plugin, you will see
that it has not seen any substantial changes since its
original conception in 2003.

molfile plugins in VMD are very much a user supported
feature. if nobody takes care to improve them, they will
stay the way they are. this is specifically true for
plugins that read file formats that the primary developer(s)
don't use. this is something that everybody should keep
in mind when asking for getting support for a new or
changed file format.

cheers,
    axel.

> Thanks,
> Ignacio
>
>
>
> 

-- 
Dr. Axel Kohlmeyer  akohlmey_at_gmail.com 
Institute for Computational Molecular Science
College of Science and Technology
Temple University, Philadelphia PA, USA.